Is college DRTG a great predictor of a player's defensive potential in the pro's

[Peoriabird]

Here is a list of the players with the best DRtg since the 2009-10 season and there are some surprises

Karl Anthony Towns comes in at #3 on this list

Muscala even made the list at # 31

https://www.sports-reference.com/cbb/leaders/def-rtg-player-season.html

I need help analyzing DRtg...What is a good rating and what is bad? And does it translate from college to the pros?

[Peoriabird]

15 minutes ago, Peoriabird said:

Here is a list of the players with the best DRtg since the 2009-10 season and there are some surprises

Karl Anthony Towns comes in at #3 on this list

Muscala even made the list at # 31

https://www.sports-reference.com/cbb/leaders/def-rtg-player-season.html

I need help analyzing DRtg...What is a good rating and what is bad? And does it translate from college to the pros?

For example Markelle Fultz had a defensive rating of 110 in college but is 7th in the NBA in DRtg

Karl Antony Towns was 3rd best in DRtg in college since 2009-10 but is not in the top 200 in Drtg in the NBA

[AHF]

On the what is good, what is bad question, DRtg is a formula trying to estimate how many points the team will give up on the floor with that player so a low number is good and a high number is bad.

It is a stat that needs a lot of context around it before you can use it for too much.  The DRtg stat is driven a lot by team performance so people can have inflated DRtgs when they are part of a great defense and unappealing ones when they are part of a bad one even if they are the weak link in the first case or the only effective defender in the second case.  Everyone on that 2014-15 UK team looks like a defensive stud because of this.  Andre Roberson's numbers go up and down with his team's performance (worse than everyone on the UK team when Colorado had an embarrassing defense and then clumped in with mid-UK 2014-15 starters once his team dropped opponents' FG% from 257th best to 522nd best during his time there.  Had he played on UK instead of Colorado he would have had exceptional numbers every year.

College numbers also tend to be less reliable given the smaller sample size combined with the wildly differing levels of competition from game to game.  (Fultz has a crazy SSS right now in the NBA making his DRtg this season completely worthless).

KAT is a good example where he played on a team with a very strong defense, had a boosted block rate due to WCS's matching up on the strongest offensive frontcourt player which gave him freedom to look for blocks more than WCS, and his physical gifts and the lower talent level in college mitigated some of his defensive weaknesses at that level.

[Peoriabird]

So how are people evaluating how good or bad Bagley will be as an NBA defender and concluding that he will project out to be much wore than the average freshman player entering the draft and Jaren Jackson will be so much better than the average freshman?

What measure are they using?

[AHF]

30 minutes ago, Peoriabird said:

So how are people evaluating how good or bad Bagley will be as an NBA defender and concluding that it will project out to be much wore than the average freshman player entering the draft and Jaren Jackson will be so much better than the average freshman?

What measure are they using?

Since Bagley and Carter both play for Duke, it is easier to compare the two of them and Carter looks like the better defender statistically (roughly 5 points better on DRtg) as well as with the eye test.  Scouts have noted Bagley's struggles on defense and the schematic adjustments that Coach K made to cover for him.  For example, despite being nearly 7 feet tall Bagley averaged 1/3 of the blocks that Carter did and put up a pretty pitiful 1.0 block per 40 minutes.  JJJ by comparison had the best DRtg on the team (so he stood out in a positive way when contrasted to his teammates) and his 5.5 bp40 is outstanding.  Given JJJ's length (a plus for his projected defense) and Bagley's (a negative for his projected defense) and it is really easy to project JJJ out as a much better defensive prospect.

Comparing them using less team oriented stats:

Bagley 1.4% STL%, 2.6% BLK%, 21.3% DRB%, 0.090 DWS/48 

Carter 1.7% STL%, 7.6% BLK%, 23.2% DRB%, 0.111 DWS/48

JJJ - 1.6% STL%, 14.3% BLK%, 19.7% DRB%, 0.138 DWS/48

[Peoriabird]

2 minutes ago, AHF said:

Since Bagley and Carter both play for Duke, it is easier to compare the two of them and Carter looks like the better defender statistically (roughly 5 points better on DRtg) as well as with the eye test.  Scouts have noted Bagley's struggles on defense and the schematic adjustments that Coach K made to cover for him.  For example, despite being nearly 7 feet tall Bagley averaged 1/3 of the blocks that Carter did and put up a pretty pitiful 1.0 block per 40 minutes.  JJJ by comparison had the best DRtg on the team (so he stood out in a positive way when contrasted to his teammates) and his 5.5 bp40 is outstanding.  Given JJJ's length (a plus for his projected defense) and Bagley's (a negative for his projected defense) and it is really easy to project JJJ out as a much better defensive prospect.

Comparing them using less team oriented stats:

Bagley 1.4% STL%, 2.6% BLK%, 21.3% DRB%, 0.090 DWS/48 

Carter 1.7% STL%, 7.6% BLK%, 23.2% DRB%, 0.111 DWS/48

JJJ - 1.6% STL%, 14.3% BLK%, 19.7% DRB%, 0.138 DWS/48

Ok so now I have to research Stl% Blk % DRB % DWS/48 to see whether that translates to NBA

[AHF]

2 minutes ago, Peoriabird said:

Ok so now I have to research Stl% Blk % DRB % DWS/48 to see whether that translates to NBA

Here is an except from a study on that which says that rebounds and blocks are extremely translatable while scoring and FG% are not very translatable.  3pt% correlates better but I'd suspect that the 3pt% is somewhat dependent on the volume with  higher volume shooters correlating more significantly (that is me spitballing, though).  (The closer to 1 the number is the higher it correlates while the closer to 0 a number is the more deviation you can expect).

http://www.basketball-statistics.com/howdoncaastatisticstranslatetothenba.html

Below are the R^2’s for the different correlations:

Points per minute: 0.3405
Field goal attempts per minute: 0.3522
Field goal percentage: 0.3436
Three-point attempts per minute: 0.6391
Three-point percentage: 0.7941
Free throw attempts per minute: 0.286
Free throw percentage: 0.7615
Rebounds per minute: 0.8312
Assists per minute: 0.8823
Steals per minute: 0.5981
Blocks per minute: 0.9327
Turnovers per minute: 0.4535
Personal fouls per minute: 0.4447

[Peoriabird]

4 minutes ago, AHF said:

Here is an except from a study on that which says that rebounds and blocks are extremely translatable while scoring and FG% are not very translatable.  3pt% correlates better but I'd suspect that the 3pt% is somewhat dependent on the volume with  higher volume shooters correlating more significantly (that is me spitballing, though).  (The closer to 1 the number is the higher it correlates while the closer to 0 a number is the more deviation you can expect).

http://www.basketball-statistics.com/howdoncaastatisticstranslatetothenba.html

Below are the R^2’s for the different correlations:

Points per minute: 0.3405
Field goal attempts per minute: 0.3522
Field goal percentage: 0.3436
Three-point attempts per minute: 0.6391
Three-point percentage: 0.7941
Free throw attempts per minute: 0.286
Free throw percentage: 0.7615
Rebounds per minute: 0.8312
Assists per minute: 0.8823
Steals per minute: 0.5981
Blocks per minute: 0.9327
Turnovers per minute: 0.4535
Personal fouls per minute: 0.4447

Here is the problem...Freshman numbers look different than senior numbers.  For example, John Collins averaged 10 rebounds per 40 minutes as a freshman and 14 as a sophomore

[Peoriabird]

12 minutes ago, AHF said:

Here is an except from a study on that which says that rebounds and blocks are extremely translatable while scoring and FG% are not very translatable.  3pt% correlates better but I'd suspect that the 3pt% is somewhat dependent on the volume with  higher volume shooters correlating more significantly (that is me spitballing, though).  (The closer to 1 the number is the higher it correlates while the closer to 0 a number is the more deviation you can expect).

http://www.basketball-statistics.com/howdoncaastatisticstranslatetothenba.html

Below are the R^2’s for the different correlations:

Points per minute: 0.3405
Field goal attempts per minute: 0.3522
Field goal percentage: 0.3436
Three-point attempts per minute: 0.6391
Three-point percentage: 0.7941
Free throw attempts per minute: 0.286
Free throw percentage: 0.7615
Rebounds per minute: 0.8312
Assists per minute: 0.8823
Steals per minute: 0.5981
Blocks per minute: 0.9327
Turnovers per minute: 0.4535
Personal fouls per minute: 0.4447

The only stat the Bagley is vastly different than jackson is block %  Which is offset by Bagley being a better rebounder.

[AHF]

46 minutes ago, Peoriabird said:

Here is the problem...Freshman numbers look different than senior numbers.  For example, John Collins averaged 10 rebounds per 40 minutes as a freshman and 14 as a sophomore

They are basically the same age.  Bagley is a signficantly older freshman than Collins was.

39 minutes ago, Peoriabird said:

The only stat the Bagley is vastly different than jackson is block %  Which is offset by Bagley being a better rebounder.

Ummm.....no.  Lots of these have significant deviations.  Deviations of 10% or greater are bolded below and in itallics if negative and underlined if positive, deviations of 5-9.9% are in itallics if negative and underlined if positive:

If you look at % deviation by statistic from the average of these 3 you get:

Bagley -10.6% STL, -68.2% BLK, -0.5% on DRB%, -20.4% DWS/48

Carter +8.5% STL, -7% BLK, +8.4% DRB%, -1.8% DWS/48

JJJ +2.1% STL, +75.1% BLK, -7.9% DRB%, +22.1% DWS/48

 

Bagley stands out like a sore thumb among this group.  This is not a comparison against the general population so take it that he is materially worse than the others in everything but rebounding.

[JTB]

This is why you in my opinion when you are drafting high in the NBA Lottery you look for the best offensive talents based on what position or positions you need.

Defense is WAY to hard to judge and I wouldn’t draft a player with a lottery pick based on his defensive ability cause it’s completely unpredictable at the next level whereas true offensive talent shown at the college level can actually transfer over well to the NBA level.

......Wiggins went over Parker cause of defense yet when healthy Parker is more of a threat on the court than Wiggins (in my opinion). KAT went over Okafor cause of defense but he actually ended up being a way better scorer than okafor too but we didn’t see that in him while he was at Kentucky.

 

only time a team don’t need to use a lottery pick for offensive talent is if you’re the warriors or rockets. If you’re those teams you can go after defense for a guy like Mo Bamba or JJJ for example.

Edited April 2, 2018 by JTB

[AHF]

7 minutes ago, JTB said:

This is why you in my opinion when you are drafting high in the NBA Lottery you look for the best offensive talents based on what position or positions you need.

Defense is WAY to hard to judge and I wouldn’t draft a player with a lottery pick based on his defensive ability cause it’s completely unpredictable at the next level whereas true offensive talent shown at the college level can actually transfer over well to the NBA level.

......Wiggins went over Parker cause of defense yet when healthy Parker is more of a threat on the court than Wiggins (in my opinion). KAT went over Okafor cause of defense but he actually ended up being a way better scorer than okafor too but we didn’t see that in him while he was at Kentucky.

 

only time a team don’t need to use a lottery pick for offensive talent is if you’re the warriors or rockets. If you’re those teams you can go after defense for a guy like Mo Bamba or JJJ for example.

The numbers above say that defensive rebounding and shot blocking is way more predictable than scoring.  Lots of guys are examples both ways in terms of numbers dropping (Doug McDermott, Derrick Williams, Jimmer Freddette, Shabazz Muhammed, etc.) and picking up (Kawi Leonard, Paul George, Anthony Davis, KAT, etc.)

[KB21]

1 hour ago, JTB said:

This is why you in my opinion when you are drafting high in the NBA Lottery you look for the best offensive talents based on what position or positions you need.

Defense is WAY to hard to judge and I wouldn’t draft a player with a lottery pick based on his defensive ability cause it’s completely unpredictable at the next level whereas true offensive talent shown at the college level can actually transfer over well to the NBA level.

......Wiggins went over Parker cause of defense yet when healthy Parker is more of a threat on the court than Wiggins (in my opinion). KAT went over Okafor cause of defense but he actually ended up being a way better scorer than okafor too but we didn’t see that in him while he was at Kentucky.

 

only time a team don’t need to use a lottery pick for offensive talent is if you’re the warriors or rockets. If you’re those teams you can go after defense for a guy like Mo Bamba or JJJ for example.

I wouldn't be so sure of that.  With a few exceptions, bigs who post a high DBPM tend to pan out in the NBA.  The highest recent defensive BPMs among bigs belong to Anthony Davis, Karl Anthony Towns, and Nerlens Noel.  The thing with these though is that to separate why the two panned out an the other one did not, you do look at offense.  Nerlens Noel was a net zero offensively.  He had a 2.9 OBPM during his only season at Kentucky.  Anthony Davis had a 7.8 OBPM with KAT having a 6.5 OBPM.  Jaren Jackson has a 10.3 DBPM with a 5.1 OBPM.  Mohamed Bamba has a 8.8 DBPM with a 2.4 OBPM.  Bamba fits the profile of Nerlens Noel, whereas JJJ is just a level below that of Anthony Davis and KAT.

DeAndre Ayton's OBPM and DBPM are identical to Jahlil Okafor's.  It's clear he derives most of his value from offense.

[Peoriabird]

1 hour ago, AHF said:

The numbers above say that defensive rebounding and shot blocking is way more predictable than scoring.  Lots of guys are examples both ways in terms of numbers dropping (Doug McDermott, Derrick Williams, Jimmer Freddette, Shabazz Muhammed, etc.) and picking up (Kawi Leonard, Paul George, Anthony Davis, KAT, etc.)

What happened with Shelden Williams?

[KB21]

Top 5 highest BPMs for bigs who have been drafted recently/will be drafted:

1.  Anthony Davis 18.7

2.  Karl Anthony Towns 17.3

3.  Jaren Jackson Jr 15.4

4.  Joel Embiid 14.9

5.  Wendell Carter Jr 13.5

 

Edited April 3, 2018 by KB21

[TheNorthCydeRises]

3 hours ago, JTB said:

This is why you in my opinion when you are drafting high in the NBA Lottery you look for the best offensive talents based on what position or positions you need.

Defense is WAY to hard to judge and I wouldn’t draft a player with a lottery pick based on his defensive ability cause it’s completely unpredictable at the next level whereas true offensive talent shown at the college level can actually transfer over well to the NBA level.

......Wiggins went over Parker cause of defense yet when healthy Parker is more of a threat on the court than Wiggins (in my opinion). KAT went over Okafor cause of defense but he actually ended up being a way better scorer than okafor too but we didn’t see that in him while he was at Kentucky.

 

only time a team don’t need to use a lottery pick for offensive talent is if you’re the warriors or rockets. If you’re those teams you can go after defense for a guy like Mo Bamba or JJJ for example.

It's the mistake that Detroit made back in 2003, when they chose Darko over Carmelo Anthony.  That move alone probably cost the Pistons 3 more championships.

[KB21]

10 minutes ago, TheNorthCydeRises said:

It's the mistake that Detroit made back in 2003, when they chose Darko over Carmelo Anthony.  That move alone probably cost the Pistons 3 more championships.

I doubt it.  Carmelo would have never subverted his own desire to be the guy to be part of a team.  He would have been the monkey wrench in their plan.  You could make a case for Wade or Bosh, but not Anthony, IMO.   Anthony has always been a one dimensional volume shooter who has never contributed to winning.

[TheNorthCydeRises]

2 hours ago, AHF said:

The numbers above say that defensive rebounding and shot blocking is way more predictable than scoring.  Lots of guys are examples both ways in terms of numbers dropping (Doug McDermott, Derrick Williams, Jimmer Freddette, Shabazz Muhammed, etc.) and picking up (Kawi Leonard, Paul George, Anthony Davis, KAT, etc.)

The difference in those two groups of players, is that the 1st group weren't athletic enough to create their own shots and be as successful as they were in college.  The 2nd group had all the physical tools to excel in the NBA on offense and defense.

 

[TheNorthCydeRises]

15 minutes ago, KB21 said:

I doubt it.  Carmelo would have never subverted his own desire to be the guy to be part of a team.  He would have been the monkey wrench in their plan.  You could make a case for Wade or Bosh, but not Anthony, IMO.   Anthony has always been a one dimensional volume shooter who has never contributed to winning.

Forever grateful.

 

 

[Peoriabird]

1 hour ago, Peoriabird said:

What happened with Shelden Williams?

@AHFShelden Williams averaged over 3 bpg over the 4 years he was at Duke but averaged only .5 bpg in the pros.  What happened?